# Approximation of Supremum of Max-Stable Stationary Processes and Pickands Constants

Let X(t),t∈R be a stochastically continuous stationary max-stable process with Fréchet marginals Φ_α, α>0 and set M_X(T)=_t ∈ [0,T] X(t),T>0. In the light of the seminal articles [1,2], it follows that A_T=M_X(T)/T^1/α converges in distribution as T→∞ to H_Z^1/α X(1), where H_Z is the Pickands constant corresponding to the spectral process Z of X. In this contribution we derive explicit formulas for H_Z in terms of Z and show necessary and sufficient conditions for its positivity. From our analysis it follows that A_T^β,T>0 is uniformly integrable for any β∈ (0,α). Further, we discuss the dissipative Rosiński (or mixed moving maxima) representation of X. Additionally, for Brown-Resnick X we show the validity of the celebrated Slepian inequality and obtain lower bounds on the growth of supremum of Gaussian processes with stationary increments by exploiting the link between Pickands constants and Wills functional. Moreover, we derive upper bounds for supremum of centered Gaussian processes given in terms of Wills functional, and discuss the relation between Pickands and Piterbarg constants.

## Authors

• 1 publication
• 4 publications
• ### Berry-Esseen bounds of second moment estimators for Gaussian processes observed at high frequency

Let Z:={Z_t,t≥0} be a stationary Gaussian process. We study two estimato...
02/09/2021 ∙ by Soukaina Douissi, et al. ∙ 0

• ### Multivariate Max-Stable Processes and Homogeneous Functionals

Multivariate max-stable processes are important for both theoretical inv...
02/12/2021 ∙ by Enkelejd Hashorva, et al. ∙ 0

• ### On the Information Dimension of Multivariate Gaussian Processes

The authors have recently defined the Rényi information dimension rate d...
12/21/2017 ∙ by Bernhard C. Geiger, et al. ∙ 0

• ### Discrete sticky couplings of functional autoregressive processes

In this paper, we provide bounds in Wasserstein and total variation dist...
04/14/2021 ∙ by Alain Durmus, et al. ∙ 0

• ### Stochastic Process Bandits: Upper Confidence Bounds Algorithms via Generic Chaining

The paper considers the problem of global optimization in the setup of s...
02/16/2016 ∙ by Emile Contal, et al. ∙ 0

• ### Minoration via Mixed Volumes and Cover's Problem for General Channels

We propose a method for establishing lower bounds on the supremum of pro...
12/28/2020 ∙ by Jingbo Liu, et al. ∙ 0

• ### Metric mean dimension and analog compression

Wu and Verdú developed a theory of almost lossless analog compression, w...
12/02/2018 ∙ by Yonatan Gutman, et al. ∙ 0

##### This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.

## 1. Introduction

Let be a stochastically continuous stationary max-stable process with Fréchet marginals . Here max-stable means that the finite dimensional distributions (fidi’s) of are max-stable multivariate distributions, see e.g., [3, 4, 5]. In the following is a non-negative process satisfying and has spectral process , i.e., we have the de Haan representation, see e.g., [3, 6] (below means equality in law)

 (1.1) X(t)d=maxi≥1PiZ(i)(t),t∈R,

where is a Poisson point process (PPP) on with intensity independent of ’s which are independent copies of . Since we consider here only stationary max-stable processes, adapting the terminology of [4] (motivated by [7]) we shall call the spectral process a Brown-Resnick stationary process.

The assumption that is stochastically continuous implies that it has a separable and measurable version, see e.g., [8]; the same holds for , see [6]. Therefore in the following we suppose that both and are jointly measurable and separable. In the sequel we shall assume further that has locally bounded sample paths, and thus by (1.1) also has locally bounded sample paths. According to [6], this assumption is important for conditions that guarantee the existence of a dissipative Rosiński (or also called a mixed moving maxima) representation of .

By separability and the locally boundedness of the sample paths of and both and

are well-defined and finite random variables for any

. Further, by (1.1) given we have (see e.g., [9, 10])

 (1.2)

Since by measurability of , for any using Fubini theorem

 E{∫T0Zα(t)λ(dt)}=∫T0E{Zα(t)}λ(dt)=T,

with the Lebesque measure on , we have further that

 P{MX(T)∈(0,∞)}=P{MZα(T)∈(0,∞)}=1

and

 (1.3) −lnP{MX(T)≤(Tx)1/α}=E{MZα(T)}Tx,T,x>0.

The above shows that does not depend on the particular choice of the spectral process but only on . By the stationarity of it follows that

 E{MZα(T)}=E{supS≤t≤S+TZα(t)}

for any . Hence is sub-additive and thus by Fekete lemma

 (1.4) HZ:=limT→∞1TE{MZα(T)}=infT>01TE{MZα(T)}≤E{MZα([rgb]0,0,01)}<∞.

Moreover, from the above we conculde that does not depend on the particular choice of the spectral tail process but only on the stationary max-stable process . Referring to [11], is the so-called generalised Pickands constant defined with respect to a Brown-Resnick stationary process . In [12] is introduced for the log-normal process

 (1.5) Z(t) = eB(t)−σ2(t)/2,t∈R,

where

is a centered Gaussian process with stationary increments, continuous sample paths and variance function

. Taking to be a fractional Brownian motion (fBm) with self-similarity Hurst index , we get that is the classical Pickands constant, see e.g., [13, 14, 15, 11]. The only known values of are 1 and corresponding to and , respectively. In the case of Lévy processes the Pickands constant appears explicitly in [16, 17, 18, 11]. Moreover, for the discrete-time case we have that (introduced similarly as for the continuous-time, see [19]) is the extremal index of the stationary time series . In this context, it has been also studied in [20] using the spectral representation of

. A huge amount of research is dedicated to calculation and estimation of the extremal index of regularly varying time series, see e.g.,

[21, 22, 23, 24, 25, 26, 27, 28, 29, 30] and the references therein.
The main question that arises for Pickands constants , or in the discrete setup for the extremal index of a stationary max-stable time series is:
Q1: Under what conditions are these constants positive or equal to 0?
For the max-stable this question is partially answered in [11] for such that almost surely, and for being a non-negative random variable in [31]. Specifically, the positivity of has been shown under the assumption that

 (1.6) P{S(Z)<∞}=1,S(Z):=∫RZα(t)λ(dt).

In view of [6] since has locally bounded sample paths, then under (1.6) has a dissipative Rosiński representation which is equivalent with being generated by a non-singular dissipative flow, see [32, 33, 34, 6] for more details. As shown in [11, 31], if has càdlàg sample paths and (1.6) holds, then

 (1.7) HZ≥E{supt∈RZα(t)S(Z)},

with equality shown under some technical assumptions for both Gaussian and Lévy case. The investigation therein was motivated by [35, 9]. The former contribution showed that (1.7) holds with equality for an fBm. Since in (1.4) is defined as a limit, it turns out that the explicit calculation of is for general too difficult. However, if (1.7) holds with equality, then being an expectation, can be efficiently simulated. Indeed this has been successfully implemented for the classical Pickands constants in [35].
An interesting question that arises here is:
Q2: Does (1.7) hold with equality for general Brown-Resnick stationary ?
Clearly, if , then (1.4

) means the convergence in probability

 (1.8) AT:=MX(T)T1/αp→0,T→∞,

whereas when we have the convergence in distribution

 (1.9) ATd→H1/αZX(1),T→∞.

For being a symmetric -stable () stationary process with the above convergence has been shown in the seminal articles [1, 2], see the recent contributions [36, 37, 38, 39, 40, 41, 42] for related results and new developments.
The findings of [1, 2] are important for the max-stable processes too, which is already pointed out in [20] for discrete max-stable processes. Indeed, using the link between max-stable and processes established in [32] and [43] independently, it follows that when is generated by a non-singular conservative flow, which by [6] (under the assumption of locally boundedness of sample paths of ) is equivalent with then we have

 (1.10) HZ=0.

Note that (1.10) holds also when we consider the discrete case , which can be shown for instance by utilising the expression of Pickands constant (which in this case coincides with the extremal index, [19, 31]) derived in [20].
We conclude that is positive if and only if . Moreover, as shown in [20] for the discrete case, can be explicitly given in terms of the spectral functions. Hence question Q1 has a simple answer, namely if is stationary max-stable process with locally bounded sample paths, then, by [6], if and only

 (1.11) P{S(Z)<∞}>0.

Clearly, the convergence in probability in (1.8) implies that as for any and a similar implication holds for when (1.9) is satisfied. For being an random field the recent contribution [44] strengthened those convergences to that of for , i.e., showing the uniform integrability of whenever . The case that is a stationary max-stable random field is easier to deal with, see Proposition 3.13 in Section 4.
Our main interest in this contribution is the derivation of expressions for in terms of the spectral process that appears in the de Haan representation (1.1). In particular, motivated by Q2, we show that (1.7) (or a modification of it) holds with equality under (1.11) without further assumptions. Recall that so far it is only known that the inequality in (1.7) holds for having a dissipative Rosiński representation.
As already shown in [11, 31], different representations for relate to different dissipative Rosiński representations of . Therefore, our analysis is also concerned with such representations for .

Our study of Pickands constants (together with the criteria for its positivity) allows us to investigate the growth of the expectations of and as . The latter can be investigated under the further assumption of the Brown-Resnick model, i.e., when is a log-normal process. Moreover, for the Brown-Resnick model an extension of the celebrated Slepian inequality is possible, see Theorem 3.1 below.

Organisation of the paper: Our main results are displayed in Section 2 followed by discussions and some extensions displayed in Section 3. Proofs are postponed to Section 4; an Appendix concludes this contribution.

## 2. Main Results

Let be as in the Introduction being jointly measurable, separable and with locally bounded sample paths. By the measurability of we have that is a random variable in , see [8]. Write next instead of for an event with . Fixing we have the following splitting formula

 (2.1) E{MZα(T)}=E{MZα(T);S(Z)<∞}+E{MZα(T);S(Z)=∞}.

If (1.11) holds, then by [45][Lem 16] the random process defined by

 ZD(t):=Z(t)I(S(Z)=∞),t∈R

is Brown-Resnick stationary.
Since has also locally bounded sample paths and almost surely, the corresponding max-stable process is generated by a non-singular conservative flow. Moreover,

 ZC(t)=Z(t)I(S(Z)<∞)

is also a Brown-Resnick stationary process which is generated by a non-singular dissipative flow, provided that (1.11) holds. In order to omit technical details, we refer the reader to the deep contributions [34, 6] for details on conservative and dissipative parts of max-stable processes.
Consequently, by the discussions in the Introduction, condition (1.11) implies that

 (2.2) HZ=HZC>0.

Therefore, in the following we can reduce our analysis by considering only the case that , i.e., is generated by a non-singular dissipative flow. In view of [6] this is equivalent with having a dissipative Rosiński representation i.e., for some non-negative random process (called also random shape function) which is continuous in probability (we can consider here therefore to be jointly measurable and separable) and for some we have

 (2.3) X(t)d=maxi≥1PiL(i)(t−Ti),t∈R,

where is a PPP on with intensity , independent of ’s which are independent copies of .
In view of [46][Thm 4.2], equation (2.3) implies that for any random variable with density

 (2.4) Z=(c/p(N))1/αBNL

is a valid spectral process for , where is independent of and . See below Theorem 3.11 for a converse result. We state next the main result.

###### Theorem 2.1.

Let be a stochastically continuous max-stable process with de Haan representation (1.1). If has locally bounded sample paths with some spectral process satisfying condition (1.6), then there exists some jointly measurable and separable non-negative random shape function such that (2.4) holds and moreover we have

 (2.5) HZ=E{supt∈RLα(t)}E{S(L)}∈(0,∞).
###### Remark 2.2.

i) If is a stationary max-stable process, then the Pickands constant is defined by (see e.g., [11])

 limn→∞1TE{supt∈δZ∩[0,T]Zα(i)}=HδZ.

In view of [6] has a dissipative Rosiński representation if and only if . Under this condition, as in the proof of Theorem 2.1, it follows that the Pickands constant is given by

 (2.6) HδZ=E{supt∈δZLα(t)}E{δ∑t∈δZLα(t)}∈(0,δ].

ii) When has a dissipative Rosiński representation, it is possible to construct such that almost surely, see [6]. Hence by (2.5) for such random shape functions we have

 HZ=1E{S(L)}.

We shall show below that it is also possible to construct such that almost surely, and thus by (2.5) we obtain an alternative formula, namely

 HZ=E{supt∈RLα(t)}.

For simplicity we shall assume in the following that both and have càdlàg sample paths. Let be the space of càdlàg functions equipped with a metric which makes it complete and separable, see e.g., [47] for details. Let be the Borel -algebra on defined by this metric and let be a probability measure given by

 μ(A)=E{Zα(0)I(Z/Z(0)∈A)},A∈D.

Since is a Polish metric space (complete and separable), we can determine a stochastic process with càdlàg sample paths and probability law ; refer to as the spectral tail process. This terminology is in agreement with [22] since as shown in [31], defined above agrees with the spectral tail process of the stationary time series , see [48, 49] for earlier and further results.
By [31][Thm 4.1] all the fidi’s of the max-stable process are determined by . Namely, we have the following inf-argmax formula valid for ’s positive constants and ’s in (see also Lemma 5.1 in Appendix for a short derivation)

 (2.7) −lnP{X(ti)≤xi,1≤i≤n} = n∑k=11xαkP{infargmax1≤i≤n(Θα(ti−tk)xαi)=k}.

Consequently, defines and vice-versa, from we can calculate the fidi’s of by the generalised Pareto distributions of , see [31][Remark 6.4] and [50] for representations of generalised Pareto distributions.

The next result gives an explicit construction for the random shape function and confirms (1.7).

###### Theorem 2.3.

Under the setup of Theorem 2.1, if further has càdlàg sample paths and (1.11) holds, then

 (2.8) HZ=E{supt∈RΘα(t)S(Θ);S(Θ)∈(0,∞)}∈(0,∞).

Moreover, if (1.6) is valid, then has dissipative Rosiński representation (2.4) with random shape function

 (2.9) L(t)=Θ(t)(S(Θ))1/α,t∈R.

Example 1. Consider the Gaussian case with as in (1.5), where is a centered Gaussian process with stationary increments, continuous sample paths and variance function .

We can assume without loss of generality (see [4]) that . Hence almost surely and for the corresponding spectral tail process we simply have . Since for this case , then under (1.11)

 HZ=E{supt∈RZ(t)S(Z);S(Z)∈(0,∞)}∈(0,∞).

In view of [4], the following condition

 (2.10) liminft→∞σ2(t)lnt>8

implies (1.1) and thus has a dissipative Rosiński representation with random shape function . Moreover

 (2.11) HZ=E{supt∈RZ(t)S(Z)}∈(0,∞),

which has been proved in [11][Thm 2] under additional (redundant) assumptions on .

Example 2. Stationary max-stable Lévy–Brown–Resnick processes are introduced in [17], where the spectral process is constructed from two independent Lévy processes. Specifically, let be a Lévy process with Laplace exponent being finite for . Write for another independent Lévy process with Laplace exponent

 lnE{eθW−(1)}=Ψ(1−θ)−(1−θ)Ψ(1).

Then we set , and if . In view of [17] the max-stable process with unit Fréchet marginals corresponding to the spectral process is stationary.
Furthermore, by [18] the Lévy-Brown–Resnick process admits a dissipative Rosiński representation and thus Theorem 2.3 and [11][Thm 3.2] imply that (note that since almost surely then )

 (2.12) HZ=E{supt∈RZ(t)S(Z)}=k––(0,1)k––′(0,0)>0,

where is the bivariate Laplace exponent of the descending ladder process corresponding to .
If is a spectrally negative Lévy process, by [17] , which agrees with the fact that

 Ψ′(1)=k––(0,1)k––′(0,0)>0.

## 3. Discussions & Extensions

### 3.1. Slepian inequality for Brown-Resnick max-stable processes

Slepian inequality is essential in the theory of extremes and sample path properties of Gaussian and related processes. Besides it is also useful in numerous fields of mathematics including optimisation and number theory problems, see e.g., [51].
A commonly used version of Slepian inequality given for instance in [52][Thm 1.1] is as follows: If are two centered Gaussian processes, then for any we have

 E{max1≤i≤nB1(ti)}≥E{max1≤i≤nB2(ti)},

provided that for all

 (3.1) E{(B1(ti)−B1(tj))2}≥E{(B2(ti)−B2(tj))2}.

Moreover, in view of [53][Eq. 6] (applied to ) for any real-valued function

 (3.2) E{max1≤i≤neB1(ti)−f(ti)}≥E{max1≤i≤neB2(ti)−f(ti)}.

Let be max-stable processes with spectral processes . By (1.2) if are separable with locally bounded sample paths such that (3.1) holds, using (3.2) we obtain

 (3.3) P{supt∈KX1(t)>x}≥P{supt∈KX2(t)>x},∀x>0.

Max-stable processes that are constructed from log-normal Gaussian spectral processes are commonly referred to in the literature as Brown-Resnick max-stable processes, a first example is given in [7], see [4, 54, 5, 55, 56] for more details and interesting statistical models.
Both processes and are Brown-Resnick stationary, see [4, 32, 57, 58, 10] if defined by

 Zi(t)=eBi(t)−σ2i(t)/2,i=1,2,t∈R,

where are two centered Gaussian processes with stationary increments and variance functions and , respectively. Since in general is different from we cannot use the refinement of Vitale [53] to Slepian inequality stated in (3.2) to arrive at (3.3).

Our next result states the Slepian inequality for Brown-Resnick stationary processes and . Moreover, it implies a comparison criteria for the corresponding Pickands constants.

It is well-known that the law of ’s depends only on their variograms , . Therefore we can suppose without loss of generality that for .

###### Theorem 3.1.

Let be two stationary max-stable Brown-Resnick processes with spectral processes and , respectively. Suppose that and are separable with locally bounded sample paths. If further for any

 (3.4) σ1(t) ≥ σ2(t),

then for any compact set

 P{supt∈KX1(t)>x}≥P{supt∈KX2(t)>x},∀x>0

and moreover

 HZ1≥HZ2>0.
###### Remark 3.2.

i) As noted in [43], it is not known if there exists any centered Gaussian process with stationary increments and variance function such that .
ii) Under the setup of Theorem 3.1, if , then by Example 1, and consequently . This implies that cannot be generated by a non-singular conservative flow, therefore we can conclude that

iii) Theorem 3.1 extends the findings of [12][Thm 3.2].
iv) Slepian inequality has been extended to non-Gaussian setup in [52]. Using the results of the aforementioned paper, Slepian inequality for max-stable processes can also be derived when and have fidi’s satisfying the conditions of [52][Thm 3.1].

### 3.2. Wills functional of Gaussian processes and asymptotic constants

In this section, motivated by Vitale [59], we discuss Wills functional for Gaussian processes and its relation with Pickands and Piterbarg constants.
Consider as in the previous section, where is a centered separable Gaussian process with variance function and bounded sample paths. In [59] the Wills functional is defined as

 (3.5) WZ(T)=E{supt∈[0,T]Z(t)}.

This functional is important for derivation of lower bounds on supremum of Gaussian processes and random fields. Specifically, by [59][Thm 1] for any we have

 (3.6) lnWZ(T)≤E{supt∈[0,T]B(t)}<∞,

where the last inequality follows by the boundedness of sample paths, see e.g., [60][(3.1)]. Our findings on Pickands constants, when is Brown-Resnick stationary can be utilised to derive asymptotic properties of Wills functionals as well as lower bounds for . Indeed, the Pickands constant is given in terms of Wills functional as

 (3.7) limT→∞WZ(T)T=HZ∈(0,∞).

As shown in [58] is Brown-Resnick stationary if and only if has stationary increments. In view of Example 1, we have:

###### Lemma 3.3.

If is a centered separable Gaussian process with stationary increments, locally bounded sample paths and variance function , then

 (3.8) liminfT→∞[E{supt∈[0,T]B(t)}−lnT] ≥ lnHZ,

provided that (2.10) holds.

###### Remark 3.4.

Sharp bounds on the expectation of maximum of Gaussian processes on finite intervals are obtained recently in [61]; see also [62]. Our result above is of interest when considering the growth of supremum of on large increasing intervals and is proportional to for some (recall (2.10) which guarantees the positivity of ).

Wills functionals are tightly related with Borell-type bounds for tail distribution of suprema of Gaussian processes. We derive below a variant of Borell inequality accommodated to the setup of this section, see also [59][Cor 1]. For a given variance function , let in the following .

###### Proposition 3.5.

Suppose that is a centered separable Gaussian process with bounded sample paths and variance . For any such that and with we have

 (3.9) P{supt∈[0,T]B(t)>u}≤WVu(T)exp(−u22σ2T)

for any .

Next, inserting in (3.9) we get a lower bound for Wills functional (3.5), namely

 ∞>WZ(T)≥P{supt∈[0,T]B(t)>σ2T}exp(σ2T2).

In the special case of having stationary increments, utilizing properties of Pickands constants, we arrive at the following corollary to Proposition 3.5. Set below , , with a standard fractional Brownian motion with Hurst index .

###### Corollary 3.6.

Suppose that is a centered Gaussian process with stationary increments, bounded sample paths and variance function . If for , with and , then for each

 P{supt∈[0,T]B(t)>u}≤WZα⎛⎝C1/αu2/αTσ4/αT⎞⎠exp(−u22σ2T).

Note that is finite and asymptotically linear in as , by (3.7).

To this end we briefly discuss Piterbarg constants, which appear in the tail asymptotics of supremum of non-stationary Gaussian processes, see e.g., [63, 64].
Given as above, the corresponding Piterbarg constants are defined by

 (3.10) PB,f(K)=E{supt∈KeB(t)−σ2(t)/2−f(t)}

for some measurable locally bounded function and or . So far Piterbarg constants with have been reported and their finiteness for general under some restrictions on is recently shown in [65][Lem 4.5] and [66][Prop 3.1]. Clearly, the main question that arises is if is finite. For , its finiteness is shown using Piterbarg’s approach, see [63]. We establish below the finiteness of Piterbarg constants for some general by using the fact that is Brown-Resnick stationary, and moreover derive a bound in terms of Wills functional. In the following proposition we consider only the case ; scenario follows by analogous line of reasoning.

###### Proposition 3.7.

If is as in Proposition 3.6, then for any measurable function such that and , we have that and moreover

 (3.11) PB,f[(0,∞))≤infδ>0WZ(δ)∞∑i=0supi∈[0,1]e−f(iδ)<∞.

### 3.3. Dissipative and conservative stationary max-stable processes

The decomposition of max-stable processes into conservative and dissipative parts has been recently discussed in [6] under some assumptions on the spectral process ; see also the recent review [67] for processes. Previous results based on the works of Rosiński [68, 69] have investigated decompositions of sum-stable and max-stable processes; see also [43, 34, 32]. Since we consider in this contribution the de Haan representation (1.1), we shall rely on the recent findings of [6]. In [45][Lem 16] it is shown that a conservative/dissipative decomposition of corresponds to the so-called cone decomposition. In view of [6] a simple criteria for a stationary max-stable process with locally bounded sample paths to be generated by a non-singular conservative flow (abbreviate this as ” is conservative”) is , where is some spectral process of . The next two results propose conditions for as above to be conservative, or for to have a dissipative Rosiński representation. For simplicity, we consider the case that has càdlàg sample paths.

###### Proposition 3.8.

Let be stationary and max-stable process with Fréchet marginals and càdlàg sample paths. Denote by its spectral tail process. Then the following are equivalent:

1. is conservative.

2. The Pickands constant equals 0.

We state next equivalent conditions for to have a dissipative Rosiński representation.

###### Proposition 3.9.

Let be as in Proposition 3.8. The following are equivalent:

1. has representation (2.3) with some càdlàg process .

2. For any compact set we have

 P{∫Rsupu∈KZα(t+u)λ(dt)∈(0,∞)}=P{∫Rsupu∈KΘα(t+u)λ(dt)∈(0,∞)}=1.
3. .

###### Remark 3.10.

The above propositions hold also for the discrete-time setup , substituting the integral by the sum respectively. Additionally, in view of [22] and [19] it follows that

 P{limt∈Z,|t|→∞Θ(t)=0}=1

is equivalent with to have a dissipative Rosiński representation, which is also equivalent with the anti-clustering condition for introduced in [21], see also [70].

### 3.4. Rosiński and De Haan Representations

In view of [6] and the above discussions, simple conditions on or guarantee the dissipative Rosiński representation (2.3). When has such a representation, using either [6] or our results here, we can construct the random shape function explicitly. Then, applying for instance [46][Thm 4.2] we can determine spectral processes as in (2.4) by choosing some random variable with positive density being independent of . Note that in the aforementioned reference with continuous sample paths are considered. The same holds under a more general assumption that is measurable.

###### Theorem 3.11.

If is a measurable non-negative process with locally bounded sample paths such that for some , then the max-stable processes corresponding to the spectral processes determined for different independent of via (2.4) are stationary with Fréchet marginals, equal in distribution, and have dissipative Rosiński representation (2.3).

###### Remark 3.12.

The constant in the representation depends only on . Indeed, since we assume that has Fréchet distribution, by (1.2)

 (3.12) −lnP{X(0)≤1}=E{Zα(0)}=cE{∫R(BtLα)(0)λ(dt)}=cE{∫RLα(t)λ(dt)}>0.

To this end we mention an important class of max-stable stationary processes for which is deterministic.

Example 3. Let be a càdlàg function with for some . If is a max-stable process with given by (2.4), then Theorem 3.11 implies that is stationary. Applying Theorem 2.3 we have

 HZ=supt∈RLα(t).

A special case is if is the density of an random variable, and thus

 HZ=1√2πσ,

which shows that can assume all values in .

### 3.5. Growth of supremum

Below let be a separable max-stable stationary process with locally bounded sample paths and spectral process such that (1.1) holds. For any and we have for any

 E{AβT} = ∫∞0P{AT>x1/β}dx = ∫∞0(1−e−E{supt∈[0,T]Zα(t)}/(Txα/β))dx = (E{supt∈[0,T]Zα(t)}/T)β/αΓ(1−β/α),

where stands for the Euler Gamma function and we used (1.3). Consequently, we have

 (3.13) limT→∞E{AβT}=Hβ/αZΓ(1−β/α)<∞.

A direct implication of (1.8), (1.9) and (3.13) is the following proposition.

###### Proposition 3.13.

If where or is a max-stable stationary process as above, then (3.13) holds and moreover is uniformly integrable for any .

###### Remark 3.14.

For a stationary random field (3.13) has been shown in [44][Thm 3.1]. Extension of (3.13) to stationary max-stable random fields is straightforward and omitted here.

### 3.6. Max-stable random fields

Max-stable random fields can be defined exactly as the max-stable processes by simply substituting the random process by a random field . The corresponding Pickands constant is then defined by

 HZ=limmin1≤i≤dTi→∞d∏i=1T−1iE⎧⎨⎩supt∈∏di=1[0,Ti]eZ(t)⎫⎬⎭,

where again we suppose that is stationary max-stable with Fréchet marginals . Since the functional is translation invariant, i.e.,

 F(A+t)=F(A),t∈Rd,A⊂Rd

and is sub-additive in the sense that for disjoint , then by [71] we have that exists and is finite.
In view of [6], with locally bounded sample paths has a dissipative Rosiński representation if , where with the Lebesgue measure in . Our findings above can be easily extended to this setup of stationary max-stable random fields. Note in passing that for the discrete-time setup the extremal index of is calculated in terms of spectral functions in [20], see also [31] for formulas in terms of spectral tail process.
Given the importance of the Gaussian model, for simplicity we state next the result for the Brown-Resnick max-stable model.

Let be a log-normal Gaussian random field where is a centered jointly measurable and separable Gaussian random field with stationary increments and variance function . If , which in light of [4] follows if

 (3.14) liminf∥t∥→∞Var(B(h)−B(0))ln∥h∥>8d,

where stands for the Euclidean norm, then we have

 (3.15) HZ=E⎧⎨⎩supt∈RdeB(t)−σ2(t)/2∫RdeB(t)−σ2(t)/2λ(dt)⎫⎬⎭∈(0,∞).

Next, we discuss an important issue relevant for the simulation of Brown-Resnick max-stable stationary random fields, a topic of great interest for various applications, see for recent developments [72, 73, 9, 49, 74, 75]. As formulated in [75][Problem 2] for simulation of the Brown-Resnick max-stable stationary random field with spectral random field with a compact subset of , it is important to be able to minimise the functional

 E{supt∈KeB(t)−σ2(t)/2+f(t)}

for all centered Gaussian random fields with stationary increments and variance function that have the same variogram for some given continuous function .
We conclude this section with the following comparison result:

###### Proposition 3.15.

For two given separable centered Gaussian random fields with stationary increments, bounded sample paths and variance functions respectively, with a compact subset of , and for any measurable and locally bounded function , we have

 (3.16) ∞>E{supt∈KeB1(t)−σ21(t)/2+f(t)}≥E{supt∈KeB2(t)−σ22(t)/2+f(t)},

provided that .

A direct implication of (3.16) and (3.11) is that we can compare also two Piterbarg constants, namely for any measurable locally bounded function as in Proposition 3.7

 PB1,f([0,∞))≥PB2,f([0,∞)).

## 4. Proofs

Proof of Theorem 2.1 We adapt the arguments of the proof of [2][Thm 2.1] for our max-stable process. Note first that by (1.11), in view of [6] has a dissipative Rosiński representation with some process which by the construction in the aforementioned paper is stochastically continuous and locally bounded (these properties are inherited from ). Therefore a jointly measurable and separable version of which is locally bounded exists, and we shall consider this version below.
Step 1: Since is given by (2.4), and moreover is locally bounded. then by (1.4) for any we have that